Single-cell discrimination based on optical tweezers Raman spectroscopy

HongFei Ma, Yong Zhang, AnPei Ye
The ability to discriminate between single cells in a label-free and noninvasive fashion is important for the classification of cells, and for the identification of similar cells from different origins. In this paper, we present the Raman spectroscopy-based identification of different types of single cells in aqueous media, and discrimination between the same types of cells from different donors using a novel Laser Tweezers Raman Spectroscopy (LTRS) technique, which combines laser trapping and micro-Raman spectroscopy. First, we measured the spectra of individual living human erythrocytes, i.e. red blood cells, and leucocytes (U937 cancer cells). High-quality Raman spectra with low fluorescence were obtained using a home-LTRS apparatus and 20 cells were measured for each cell type. The smoothing, baseline subtraction, and normalization of the data were followed by a principal components analysis (PCA). The PCA loading plots showed that the two different types of cells could be completely separated based only on the first component (PC1) (i.e. the peaks at 1300 cm−1); the discrimination accuracy could therefore reach 100%. More than 50 spectra were taken for each erythrocyte obtained from the four healthy volunteers. The average discrimination accuracy was 84.5% for two random individuals taken from the four volunteers, according to the first and second PCs. This work demonstrates that LTRS is a powerful tool for the accurate identification and discrimination of single cells, and it has the potential to be applied for the highly sensitive identification of cells in clinical diagnosis and medical jurisprudence.
DOI